Towards improved decision support in the assessment and management of pain for people with dementia in hospital:a systematic meta-review and observational study

BACKGROUND: Pain and dementia are common in older people, and impaired cognitive abilities make it difficult for them to communicate their pain. Pain, if poorly managed, impairs health and well-being. Accurate pain assessment in this vulnerable group is challenging for hospital staff, but essential for appropriate management. Robust methods for identifying, assessing and managing pain are needed.AIMS AND OBJECTIVES: Two studies were undertaken to inform the development of a decision support tool to aid hospital staff in the recognition, assessment and management of pain. The first was a meta-review of systematic reviews of observational pain assessment instruments with three objectives: (1) to identify the tools available to assess pain in adults with dementia; (2) to identify in which settings they were used and with what patient populations; and (3) to assess their reliability, validity and clinical utility. The second was a multisite observational study in hospitals with four objectives: (1) to identify information currently used by clinicians when detecting and managing pain in patients with dementia; (2) to explore existing processes for detecting and managing pain in these patients; (3) to identify the role (actual/potential) of carers in this process; and (4) to explore the organisational context in which health professionals operate. Findings also informed development of health economics data collection forms to evaluate the implementation of a new decision support intervention in hospitals.METHODS: For the meta-review of systematic reviews, 12 databases were searched. Reviews of observational pain assessment instruments that provided psychometric data were included. Papers were quality assessed and data combined using narrative synthesis. The observational study used an ethnographic approach in 11 wards in four UK hospitals. This included non-participant observation of 31 patients, audits of patient records, semistructured interviews with 52 staff and four carers, informal conversations with staff and carers and analysis of ward documents and policies. Thematic analysis of the data was undertaken by the project team.RESULTS: Data from eight systematic reviews including 28 tools were included in the meta-review. Most tools showed moderate to good reliability, but information about validity, feasibility and clinical utility was scarce. The observational study showed complex ward cultures and routines, with variations in time spent with patients, communication patterns and management practices. Carer involvement was rare. No pain decision support tools were observed in practice. Information about pain was elicited in different ways, at different times, by different health-care staff and recorded in separate documents. Individual staff made sense of patients’ pain by creating their own ‘overall picture’ from available information.LIMITATIONS: Grey literature and non-English-language papers were excluded from the meta-review. Sample sizes in the observational study were smaller than planned owing to poor documentation of patients’ dementia diagnoses, gatekeeping by staff and difficulties in gaining consent/assent. Many patients had no or geographically distant carers, or a spouse who was too unwell and/or reluctant to participate.CONCLUSIONS: No single observational pain scale was clearly superior to any other. The traditional linear concept of pain being assessed, treated and reassessed by single individuals did not ‘fit’ with clinical reality. A new approach enabling effective communication among patients, carers and staff, centralised recording of pain-related information, and an extended range of pain management interventions is proposed [Pain And Dementia Decision Support (PADDS)]. This was not tested with users, but a follow-on study aims to codesign PADDS with carers and clinicians, then introduce education on staff/patient/carer communications and use of PADDS within a structured implementation plan. PADDS will need to be tested in differing ward contexts.FUNDING: The National Institute for Health Research Health Services and Delivery Research programme